Essays on Growth Econometrics and Endogenous Information

Abstract: On Testing the Relationship between Human Capital and Economic GrowthThis paper investigates existing methodologies for testing the so-called Nelson-Phelps approach versus the Lucas approach. The paper shows that one cannot distinguish between these approaches using the existing methodologies. The paper proposes an alternative methodology based on cointegration analysis which is able to distinguish both approaches. The application to cross-country data provides evidence for the extended Neoclassical growth model, the Lucas approach and the Nelson-Phelps approach with exogenous steady state growth rate of the technology frontier and rejects the approaches with endogenous- and semi-endogenous technological progress. A General Equilibrium Framework of Information, Expectation Formation and Decision under UncertaintyThis paper introduces an information framework which is combined with the recently developed similarity-based approach to decision- and expectation formation. The paper provides a closed and interrelated system where the variables of interest and information are simultaneously determined. The paper illustrates that the framework includes standard macroeconomic learning approaches such as rational expectations as well as statistical learning and naive expectations as special cases. A Behavioral Foundation of DiscriminationUp until now, there has been no decision-theoretical foundation for discrimination based on acceptable primitives. The reason is that with standard assumptions about knowledge and cognitive ability of economic agents, discrimination cannot arise, as is illustrated by the paper. It relaxes the above assumptions in favor of experimental evidence and proposes a framework based on case-based decision theory combined with an information gathering process. Such a framework can provide a behavioral foundation for discriminating preferences derived from the basic assumptions of ambiguity aversion and different sizes of the population groups.